• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 81 subscribers
  • Views 3657 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TMS320F28379D: SPI configuration to communicate with ET1100.

Dawid Balcerzak
Expert 1365 points
Part Number: TMS320F28379D
Other Parts Discussed in Thread: C2000WARE, CONTROLSUITE

Hello everyone.

My goal is to write functions to read and write memory by SPI in ET1100. Below hardware configuration :

- TMS320F28379D: SPI-C muxed on 100, 101 and 102 GPIO, GPIO 103 as CS active low, 200MHz, LSPCLK 50 MHz.

- ET1100: SPI mode 3, SPI sel polarity: active low, data output sample: normal.

I was trying to configure SPI step by step as it's desribed in 379D' manual. So below I placed my configuration functions:

void spi_fifo_init()
{
    SpicRegs.SPIFFTX.all = 0xE040;
    SpicRegs.SPIFFRX.all = 0x2044;
    SpicRegs.SPIFFCT.all = 0x0;

    InitSpiC();
}

void InitSpiC(void)
{
    // Initialize SPI-C

    // Enable master (0 == slave, 1 == master)
    SpicRegs.SPICTL.bit.MASTER_SLAVE = 1;

    // Set reset low before configuration changes
    SpicRegs.SPICCR.bit.SPISWRESET = 0;

    // Clock polarity (0 == rising, 1 == falling)
    SpicRegs.SPICCR.bit.CLKPOLARITY = 0;

    // Clock phase (0 == normal, 1 == delayed)
    SpicRegs.SPICTL.bit.CLK_PHASE = 0;

    // Set the baud rate
    SpicRegs.SPIBRR.bit.SPI_BIT_RATE = 99;

    // 8-bit character
    SpicRegs.SPICCR.bit.SPICHAR = (8-1);

    // Enable transmission (Talk)
    SpicRegs.SPICTL.bit.TALK = 1;

    // SPI interrupts are disabled
    SpicRegs.SPICTL.bit.SPIINTENA = 0;

    // Set FREE bit
    // Halting on a breakpoint will not halt the SPI
    SpicRegs.SPIPRI.bit.FREE = 1;

    // Release the SPI from reset
    SpicRegs.SPICCR.bit.SPISWRESET = 1;
}

void InitSpicGpio()
{
   EALLOW;
    // Enable internal pull-up for the selected pins

    //GpioCtrlRegs.GPDPUD.bit.GPIO100 = 0;  // Enable pull-up on GPIO100 (SPISIMOA)
    GpioCtrlRegs.GPDDIR.bit.GPIO100 = 1;    // mosi as output

    //GpioCtrlRegs.GPDPUD.bit.GPIO101 = 0;  // Enable pull-up on GPIO101 (SPISOMIA)
    GpioCtrlRegs.GPDDIR.bit.GPIO101 = 0;  // miso as output

    GpioCtrlRegs.GPDPUD.bit.GPIO102 = 0;  // Enable pull-up on GPIO102 (SPICLKA)
    GpioCtrlRegs.GPDDIR.bit.GPIO102 = 1;    // clk as output

    GpioCtrlRegs.GPDDIR.bit.GPIO103 = 1;    // cs as output
    GpioCtrlRegs.GPDPUD.bit.GPIO103 = 0;  // Enable pull-up on GPIO103 (SPISTEA)

    //
    // Set qualification for selected pins to asynch only
    //
    // This will select asynch (no qualification) for the selected pins.
    // Comment out other unwanted lines.
    //
    GpioCtrlRegs.GPDQSEL1.bit.GPIO100 = 3; // Asynch input GPIO16 (SPISIMOA)
    GpioCtrlRegs.GPDQSEL1.bit.GPIO101 = 3; // Asynch input GPIO17 (SPISOMIA)
    GpioCtrlRegs.GPDQSEL1.bit.GPIO102 = 3; // Asynch input GPIO18 (SPICLKA)
    GpioCtrlRegs.GPDQSEL1.bit.GPIO103 = 3; // Asynch input GPIO19 (SPISTEA)

    //Configure SPI-C pins using GPIO regs
    GpioCtrlRegs.GPDGMUX1.bit.GPIO100 = 1; // Configure GPIO16 as SPISIMOA
    GpioCtrlRegs.GPDGMUX1.bit.GPIO101 = 1; // Configure GPIO17 as SPISOMIA
    GpioCtrlRegs.GPDGMUX1.bit.GPIO102 = 1; // Configure GPIO18 as SPICLKA
    GpioCtrlRegs.GPDGMUX1.bit.GPIO103 = 1; // Configure GPIO19 as SPISTEA

    GpioCtrlRegs.GPDMUX1.bit.GPIO100 = 2; // Configure GPIO16 as SPISIMOA
    GpioCtrlRegs.GPDMUX1.bit.GPIO101 = 2; // Configure GPIO17 as SPISOMIA
    GpioCtrlRegs.GPDMUX1.bit.GPIO102 = 2; // Configure GPIO18 as SPICLKA
    GpioCtrlRegs.GPDMUX1.bit.GPIO103 = 2; // Configure GPIO19 as SPISTEA

    EDIS;
}

Could you tell me if my SPI configuration is genuine? I am not sure because I am trying to read 0x0 address from ET1100 (information about type of a device) and nothing is happening on txbuffer. Below my test code:

    unsigned int byte0 = 0x0;
    unsigned int byte1 = 0b00000011;
    unsigned int byte3 = 0xFF;

    GpioDataRegs.GPDCLEAR.bit.GPIO103 = 1;
    asm(" NOP");    //need 15ns delay
    asm(" NOP");
    asm(" NOP");
    asm(" NOP");

    SpicRegs.SPIDAT = (byte0 << 8); // sending first address byte
    while(SpicRegs.SPISTS.bit.BUFFULL_FLAG);

    SpicRegs.SPIDAT = (byte1 << 8); // sending second address byte
    while(SpicRegs.SPISTS.bit.BUFFULL_FLAG);
    
    SpicRegs.SPIDAT = (byte3 << 8); // sending wait state byte
    while(SpicRegs.SPISTS.bit.BUFFULL_FLAG);

    SpicRegs.SPIDAT = (byte3 << 8); // sending read termination byte
    while(SpicRegs.SPISTS.bit.BUFFULL_FLAG);

Any tip will be very appreciated.

ET1100 datasheet: download.beckhoff.com/.../ethercat_et1100_datasheet_v1i9.pdf

  • 0
    Expert 1365 points

    Could you tell me if my SPI configuration is genuine? I am not sure because I am trying to read 0x0 address from ET1100 (information about type of a device) and nothing is happening on rxbuffer. I posted my test code in previous post. I am trying to use 2 byte  Address modes for Read access with Wait state byte.

    Here are tables from ET110 datasheet:

    And example SPI routine:

  • 0 in reply to Dawid Balcerzak
    TI__Genius 15260 points
    Hi Dawid,

    Have you tried enabling loopback mode on our SPI to see if you are correctly receiving the data you are attemping to send? We have examples for this in C2000Ware available for reference.

    Have you probed the signals between the devices with an oscilloscope?

    Regards,
    Kris
  • 0 in reply to Kris Parrent
    Expert 1365 points

    Hi Kris,

    of course I tried with loopback and example from TI, from controlSUITE specifically. And sent data are correct.

    In my case it is difficult to debug signals with an oscilloscope because I am using J9 socket in LaunchPadXL-F28379D to link microcontroller with EtherCAT adapter.

  • 0 in reply to Dawid Balcerzak
    TI__Genius 15260 points
    Dawid,

    I understand. Are there other SPI pins available on your launchpad just to confirm the TX waveforms are as you expect? You could use a different pinout for the same SPI or copy the exact same configuration to a different SPI (SPI-B for example).

    Regards,
    Kris
  • 0 in reply to Kris Parrent
    Expert 1365 points

    I just noticed that something is coming to rxbuffer but it is not expected value. I think that I have a problem with algorhitm. Do you know how send orders properly?

  • 0 in reply to Dawid Balcerzak
    TI__Genius 15260 points
    Dawid,

    Can you post the received vs expected data? If it is bit-shifted it may be a result of your polarity / phase settings. I have not worked with the Beckhoff part. I will review the datasheet to see if anything jumps out for your configuration.

    Regards,
    Kris
  • 0 in reply to Kris Parrent
    Expert 1365 points

    I was wrong, I have no data on MISO.

    But I scoped signals on LaunchPad and everything is ok, I have clk signal and MOSI data. To set GPIO mux parameters I used PinMux tool from TI cloud just to make sure that configuration is correct. In my project I just set GPDDIR register

    I would like to read first register on ET1100, its address is 0x0000. Data I send:

    //spi_xmit(sdata);
        // 0 byte: 0b00000000, 1 byte: 0b00000011, 2 byte: 0b11111111, 3 byte: 0b11111111
        GpioDataRegs.GPDCLEAR.bit.GPIO103 = 1;
        DELAY_US(1);
        spi_xmit((0b00000000 << 8) | 0b00000011);
        while(SpicRegs.SPIFFRX.bit.RXFFST !=1) { }

        rdata[0] = SpicRegs.SPIRXBUF;

        DELAY_US(1);
        spi_xmit((0b11111111 << 8) | 0b11111111);
        while(SpicRegs.SPIFFRX.bit.RXFFST !=1) { }

        rdata[1] = SpicRegs.SPIRXBUF;